The Pulse
NYC Council Secures AI Lab Testimony, Subpoenas SpaceXAI
Meta confirmed voluntarily, while Anthropic, OpenAI and Google agreed to testify after subpoena warnings. The Council also plans to consider AI safety proposals at its October 5 hearing and subpoenaed SpaceXAI to compel its appearance.
AI.info Team · Written with AI ·
Three Labs Agreed After Subpoena Warnings; Meta Volunteered
On September 28, New York City Council Speaker Julie Menin announced that Anthropic, OpenAI, Google and Meta had agreed to send representatives to testify under oath at a Council hearing on artificial intelligence. The announcement came ahead of a Committee of the Whole hearing scheduled for October 5, where all 51 Council members are set to examine AI risks and proposed city safeguards.
Meta had confirmed a senior leader’s attendance earlier, in response to the Council’s initial request. According to the Council, OpenAI and Google agreed to appear on September 27, and Anthropic confirmed late that night, after the Council warned the companies they could face subpoenas. Menin said the companies should answer questions and advise on legislation in public.
“Given the high stakes, these firms owe it to the public to come before the Council, answer our questions, and provide input on our proposed legislation under oath.”
Julie Menin, New York City Council speaker
SpaceXAI Was the Only Firm Still Uncommitted
Elon Musk’s SpaceXAI had not responded to the Council’s inquiry by September 28. Menin issued a subpoena that day to compel the company to appear, making its status different from the four firms that had agreed to testify.
The Council cited Section 29 of the New York City Charter and Council Rule 7.150 as authority to investigate city matters, require attendance and take testimony under oath. If SpaceXAI does not comply, the Council said it may ask New York State Supreme Court to enforce the subpoena.
The Proposals Reach Beyond AI Companies
The hearing’s agenda also includes bills addressing safety incidents, consumer harms and oversight of AI systems. One proposal, Introduction 2602, would require third-party validation before a business markets, sells or deploys an AI system in New York City. It would also require a human-operated shutoff, with both the business and validator facing a $25,000 penalty for each instance of deployment without validation or with falsified validation.
Other proposals would let whistleblowers receive part of fines recovered from AI companies that violate applicable laws, and create a private right of action for people harmed by foreseeable misuse or circumvention of AI safety controls. Under the proposal, liability would require that the harm was foreseeable to the company, the company failed to implement reasonable safeguards, and a third party exploited that failure to cause harm. A separate bill would require a covered contractor or the agency that awarded its contract to notify the Office of Cyber Command in writing within 24 hours after becoming aware during the contract term of an AI safety incident relating to the covered contract. The office would then have 24 hours to publicly disclose a reported incident. These are proposals for Council consideration, not enacted rules.
The package also includes a plan for city agencies to respond to AI-related disruptions to municipal systems and public safety, along with measures on chatbot privacy and disclosures about AI tools. The Council said it had already enacted city laws concerning agency use of AI and algorithmic systems. The Council’s bill summary lists the measures slated for discussion.
October 5 Hearing Puts Testimony and Bills Together
The Council scheduled the October 5 session as a Committee of the Whole hearing, bringing all 51 members together to hear from company representatives and national experts in AI safety, technology and consumer protection. The Council says the session will examine risks to New Yorkers as well as possible legislative responses.
The September 28 announcement establishes who the Council says agreed to testify and why it subpoenaed SpaceXAI; it does not report the hearing’s outcome. Its central distinction is procedural: four firms committed to appear, while the Council invoked its subpoena authority to compel a fifth.